Download - Evaluation of adhesion capacity, cell surface traits and immunomodulatory activity of presumptive probiotic Lactobacillus strains

International Journal of Food Microbiology 140 (2010) 154–163

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j ourna l homepage: www.e lsev ie r.com/ locate / i j foodmicro

Evaluation of adhesion capacity, cell surface traits and immunomodulatory activity ofpresumptive probiotic Lactobacillus strains

Charalambos Kotzamanidis a, Andreas Kourelis a, Evanthia Litopoulou-Tzanetaki b,Nikolaos Tzanetakis b, Minas Yiangou a,⁎a Department of Genetics, Development & Molecular Biology, Biology School, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greeceb Laboratory of Food Microbiology and Hygiene, Faculty of Agriculture, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece

⁎ Corresponding author. Tel.: +30 2310998333; fax:E-mail address: [email protected] (M. Yiangou).

0168-1605/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.ijfoodmicro.2010.04.004

a b s t r a c t
a r t i c l e i n f o
Article history:Received 3 February 2010Received in revised form 29 March 2010Accepted 1 April 2010

Keywords:AdhessionImmunomodulationLactobacillusProbioticsAggregation

Twelve lactobacilli previously isolated from newborn infants' gastrointestinal tract and Feta cheese werefurther characterized by pulse field gel eletrophoresis (PFGE). All strains exhibited distinct PFGE genotypicpatterns with the exception of DC421 and DC423 strains possessing identical patterns. The strains DC421,2035 and 2012 were found to posses certain cell surface traits such as hydrophobicity, autoaggregation and/or high adhesive capacity suggesting potential immunomodulatory activity. However, application of thedorsal mouse air pouch system revealed that only the DC421, DC429 and 2035 strains exhibited strongimmunostimulatory activity such as increased chemotaxis of polymorphonuclear (PMN) cells in associationwith increased phagocytosis and cytokine production. The same strains also induced immunomodulatoryactivity in the gut associated lymphoid tissue in mice in the absence of any inflammatory response. Allstrains induced IgA production while reduced TNFα production by small intestine cells. The strains DC421and DC429 exerted their effect on the intestine through Toll-like receptor TLR2/TLR4/TLR9 mediatedsignalling events leading to secretion of a certain profile of cytokines in which gamma interferon (IFN-γ),interleukin (IL)-5, IL-6 and IL-10 are included. The strain 2035 induced similar cytokine profile through thesynergy of TLR2/TLR4. This study further supports the eligibility of the air pouch model to discriminatepresumptive probiotic Lactobacillus strains exhibiting immunostimulatory activity in the gut. Furthermore,evidence is provided that the cell surface traits examined may not be the only criteria but an alternative andimportant component of a complex mechanism that enables a microorganism to interact with the host gut toexert its immunoregulatory activity.

+30 2310998319.

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Lactobacillus strains isolated from the human intestinal tract andfermented dairy products are generally regarded as safe (GRAS) andhave been extensively exploited for their probiotic properties (Crosset al., 2002; Puertollano et al., 2008). Probiotics are living micro-organisms which upon ingestion in adequate amounts confer healthbenefits to the host (FAO/WHO, 2002). However, “there is a goodevidence that specific strains of probiotics are safe for human useand able to confer some health benefits on the host, but thesebenefits cannot be extrapolated to other strains without experimen-tation” (FAO/WHO, 2001). Survival and colonization in the digestivetract as well as bacterial antagonism are considered critical propertiesto ensure health benefit in the host. Of particular interest amongprobiotic properties of lactobacilli is their ability to modulate the

innate proinflammatory and antinflammatory immunity and/or adap-tive host immune responses, especially the Th1/Th2 and Treg balance(Wittig and Zeitz, 2003; Mohamadzadeh et al., 2005). Lactobacillusspecies were also shown to exert immuno-enhanching effects on thephagocytic activity of monocytes and PMN cells as well as on IgA,chemokine and cytokine production (Vaarala, 2003). These responsesaremediated in part by the recognition ofmicrobe-associatedmolecularpatterns (MAMPs) through the Toll-like receptors (TLRs) present onintestinal epithelial (IECs) as well as on, dendritic andmacrophage cells(Neish, 2002).

In the process of screening for new probiotic strains there are notclearly established bacterial phenotypic markers which could be usedfor the prediction of the immunomodulatory capacity of lactobacilli(Voltan et al., 2007). It has been suggested that adhesion ability to theintestinal epithelium is one of the most important characteristics oflactobacilli as well as one of the main criteria for selecting probioticstrains (Ouwehand et al., 1999). It has been demonstrated that thisability as well as other phenotypic traits of lactobacilli such ashydrophobicity (Pelletier et al., 1997), autoaggregation ability (Voltan

mailto:[email protected]

http://dx.doi.org/10.1016/j.ijfoodmicro.2010.04.004

http://www.sciencedirect.com/science/journal/01681605

155C. Kotzamanidis et al. / International Journal of Food Microbiology 140 (2010) 154–163

et al., 2007), the presence of cell surface-layer proteins (S-layerproteins) (Frece et al., 2005a) and the production of exopolysaccharides(EPS) (Sengul et al., 2006) are involved in the differentialmodulation ofthe host immune response.

In previous studies several Lactobacillus and yeast strains wereisolated and characterized from newborn infants' gastrointestinaltract and Feta cheese (Tzanetakis and Litopoulou-Tzanetaki, 1992;Xanthopoulos et al., 1999, Kourelis et al., 2010b). The Lactobacillusstrains belonging to five different species, Lactobacillus plantarum,Lactobacillus paracasei subsp. paracasei, Lactobacillus gasseri, Lactoba-cillus rhamnosus and Lactobacillus reuteri were found to exhibitbiotechnologically important characteristics and potential probioticstrains were selected on the basis of in vitro selection criteria such asbile salt tolerance, ability to grow at low pH and in human gastricjuice, antibacterial activity and cholesterol reduction (Xanthopouloset al., 2000a, b).

The six-day-old epithelium enclosed air pouch is formed aftersubcutaneous injection of sterile air and is developed by epithelial,phagocytic and fibroblast-like lining cells followed by an organizedvasculature that acts as a mechanical barrier (Sedgwick et al., 1985;Coates and McColl, 2001). Administration of antigen in the air pouchactivates the rapid recruitment of PMN cells while the immuneresponse products such as cytokines and chemokines are retained inthe air pouch system, enhancing early inflammatory responses noteasily detectable in systemic immunity. Bacterial administration inthe air pouch involves their interaction with the barrier formingresident macrophages and epithelial cells (Yam et al., 2008). Wehave recently demonstrated that the dorsal air pouch systemprovides the environment that permits the examination and ofinnate immune reactions in the gut and thus may be used as a modelto rapidly and efficiently discriminate potential probiotic Lactobacil-lus or yeast strains. This system was used to discriminate severalL. paracasei strains exhibiting increased chemotactic activity, PMNphagocytosis and cytokine production in both the air pouch and thegut mucosa (Kourelis et al., 2010a, c).

The aim of this study was to further investigate the in vitro pro-biotic selection criteria of 12 lactobacilli isolated from humangastrointestinal tract and Feta cheese by determining adhesion capacityand cell surface traits such as autoaggregation ability, hydrophobicity,the presence of S-layer and the production of EPS and to evaluate theirimmunomodulatory activity in both the dorsal air pouch and gutmucosa. We demonstrated the immunoregulatory activity of 3Lactobacillus strains in both the dorsal air pouch and the intestinalmucosa. Our study also provides evidence that the cell surface traitsexaminedmay not be the only criteria but an alternative and importantcomponent of a complex mechanism that enables a microorganism tointeract with the host gut to exert its immunoregulatory activity.

2. Materials and methods

2.1. Bacterial strains

Lactobacillus isolates (Table 1) were obtained from the collectionof the Laboratory of Food Microbiology and Hygiene, AristotleUniversity of Thessaloniki and were isolated from Feta cheese andinfants' gastrointestinal tract (Tzanetakis and Litopoulou-Tzanetaki,1992; Xanthopoulos et al., 1999). The L. reuteri (DC420, DC421 andDC423), L. paracasei subsp. paracasei (DC411 and DC415), L.rhamnosus (DC424 and DC429), L. gasseri (DC418 and DC422) andthe L. plantarum (2012, 2022 and 2035) strains may be considered aspresumptive probiotics on the basis of their capacity to tolerate bilesalt, to grow at low pH and in human gastric juice, to exhibitantibacterial activity and cholesterol reduction (Xanthopoulos et al.,2000a, b Tzanetakis and Litopoulou-Tzanetaki, unpublished data). Theprobiotic strain L. acidophilus NCFB 1748 whose immunostimulatory

effects in mice had been previously described (Nerstedt et al., 2007),was also included as control.

2.2. Growth conditions and preparation of the bacterial strains

All lactobacilli were subcultered three times prior to experimentaluse. For pulsed field gel electrophoresis (PFGE), in vitro assays and oraladministration in mice or injection into the air pouch, lactobacilliwere grown anaerobically for 18 h at 37 °C in MRS broth (Oxoid,Basingstoke, UK). The cells were harvested by centrifugation at3000×g for 10 min, washed two times and finally resuspended at theappropriate concentration in phosphate buffered saline (PBS). Thenumber and viability of lactobacilli were determined by culturing onMRS plates.

2.3. PFGE analysis

Genomic DNA of Lactobacillus isolates was prepared in agaroseplugs and digested with the enzyme SfiI (New England Biolabs, Inc.,USA) as previously described (Jacobsen et al., 1999). Restrictionendonuclease digestion patterns of chromosomal DNA were analyzedusing the Rotaphor type V electrophoresis unit (Biometra Gmbh,Germany). DNA fragments were then separated as previouslydescribed (Kotzamanidis et al., 2009) in 1% (w/v) PF Certified Agarose(Bio-Rad Laboratories, Inc., USA) gel in 0.25×TBE buffer, at a constantvoltage of 180 V with pulsed time ramped from 2 to 15 s linear, fieldangle 110° to 120° linear, at 22 °C for 22 h. The agarose gels were thenstainedwith ethidium bromide (0.5 μg/ml) while the Low Range PFGEMarker (2.03–194.0kbp; Bio-Rad Laboratories, Inc., USA) served asmolecular size marker and normalization reference. Calculation of thesimilarity of the band profile and grouping of the PFGE patternswere performed with Fingerprinting II Software version 3.0 (Bio-RadLaboratories, Inc., USA) by using the Pearson correlation coefficientand the unweighted pair group method using arithmetic averages(UPGMA) cluster analysis. To determine the minimum percentagesimilarity necessary for strain discrimination in PFGE analysis,reproducibility studies were carried out as previously described(Zdragas et al., 2008).

2.4. Adhesion assay

The adhesive capacity of lactobacilli included in this study wasexamined using Caco-2 intestinal epithelial cell line as previouslydescribed (Chauviere et al., 1992). Briefly, Caco-2 cells were routinelygrown in Eagle's minimal essential medium (EMEM) supplementedwith 10% (v/v) foetal bovine serum, 2 mM L-glutamine, 1% (w/v) nonessential amino acids, 100 μg/ml streptomycin, 100 IU/ml penicillin(Sigma, St Louis, MO, USA) and incubated at 37 °C in an atmospherewith 5% carbon dioxide. For adhesion assays, monolayers wereprepared on glass coverslips in 12 well tissue culture plates. Caco-2cells were seeded at a concentration of 4×104cells/cm2 and wereincubated at 37 °C in a 5% CO2 incubator. At post-confluence, 108 cellsof overnight Lactobacillus cultures were added over the monolayer ofCaco-2 cells. The plates were incubated for 90 min at 37 °C under 5%CO2 atmosphere followed by washing twice with PBS to releaseunbound bacteria, fixing with methanol and finally staining withGiemsa solution. The adherent lactobacilli were counted by micro-scopic examination under oil immersion of 20 random microscopicfields per each glass coverslip monolayer. Each adhesion assay wasperformed in triplicate with cells from three successive passages.Adhesive capacity of bacterial strains was evaluated on arbitrary scale.Bacterial strains were scored as non-adhesive when fewer than fivebacteria adhered to 100 cells and adhesive when six to 40 bacteriaadhered to 100 cells, while strong adhesion occurred with more than40 bacteria adhering to 100 cells (Del Re et al., 2000).

Table 1Adhesion capacity and cell surface characteristics of Lactobacillus strains.

Species Strains Source Hydrophobicity (%) S-layer proteins Exopolysaccharides Autoaggregation (%) Adherence to Caco-2⁎

L. paracasei DC411 Infant's intestine 1.4±0.1a − − 17.5±0.3a +L. paracasei DC415 1.4±0.2a − − 29.6±±1.4b +L. reuteri DC420 2.3±0.2a − − 26.3±2.9b −L. reuteri DC421 43.3±1.6b − − 62.6±1.2c ++L. reuteri DC423 3.5±0.3a − − 15.6±0.9a −L. gasseri DC418 2.1±0.2a − − 14.6±1.8a −L. gasseri DC422 4.6±0.3a − − 19.3±1.2a −L. rhamnosus DC424 3.4±0.3a − − 14.7±1.2a −L. rhamnosus DC429 1.8±0.5a − − 26.7±0.9b −L. plantarum 2012 Feta cheese 1.6±0.2a − + 13.7±0.9a −L. plantarum 2022 13.5±0.3c − − 15.7±1.5a +L. plantarum 2035 61.3±2.3d − − 44.3±1.9d ++L. acidophilus NCFB 1748 Human intestine 22.1±2.6e + − 19.0±1.5a −

Values are means±SEM of three independent experiments conducted in triplicates. Mean values with no common superscript letters differ significantly, Pb0.05.⁎ −, non-adhesive strains; +, adhesive strains; ++, strongly adhesive strains.

156 C. Kotzamanidis et al. / International Journal of Food Microbiology 140 (2010) 154–163

2.5. Autoaggregation assay

Autoaggregation assay was performed as described previously(Kos et al., 2003) with the following modifications. Briefly, 4 mlcontaining 108CFU/ml bacterial cells were resuspended by vortexingfor 10 s and incubated for 5 h at room temperature. At 1 h intervals,0.1 ml of the upper suspension were carefully removed, transferred toanother tube containing 3.9 ml of PBS, and the A600 was measured.The autoaggregation percentage was expressed as a function of timeuntil it was constant, using the formula: 1−(At /A0)×100, where At

represents the absorbance at time t=1, 2, 3, 4 or 5 h and A0 theabsorbance at t=0.

2.6. Cell surface hydrophobicity

Bacterial cell surface hydrophobicity was assessed by measuringmicrobial adhesion to hydrocarbons (MATH) as previously described(Crow et al., 1995). Cells at the stationary phase were washed twice inPBS and finally resuspended in 3 ml of 0.1 M KNO3 containing about108CFU/ml of bacteria and the absorbance was measured at 600 nm(A0). One milliliter of xylene was then added to the cell suspensionto form a two-phase system. After a 10-min pre-incubation at roomtemperature, the two-phase system was mixed by vortexing for2 min. Then, the water and xylene phases were allowed to separate byincubation for 20 min at room temperature. The aqueous phase wascarefully removed and its absorbance at 600 nm (A1) was measured.The percentage of cell surface hydrophobicity (H%) was calculatedusing the formula: H%=(1−A1/A0)×100.

2.7. Screening for EPS-producing lactobacilli

EPS production from Lactobacillus isolates was tested according tothe method described by Van Geel-Schutte et al. (1998) withmodifications. Briefly, Lactobacillus cultures were grown in conicalflasks containing 200 ml MRS broth supplemented with 2% (w/v)glucose, at 37 °C for 3 days. Bacterial cells were removed by centrifu-gation at 6000×g for 20 min and two volumes of 95% (v/v) cold ethanolwere added to one volume of culture supernatants for EPS precipitation.Precipitates were recovered by filtration under vacuum, dried at 60 °Cand their weight was measured to determine the amount of the EPSproduced.

2.8. Extraction of S-layer proteins and SDS-PAGE

S-layer extraction was applied on overnight cultures of lactobacillicells as previously described by Garrote et al. (2004). Briefly, 200 ml oflactobacilli cultures were centrifuged followed by washing twice with

PBS. For S-layer extraction, cell pellets were suspended in 20 ml of 5 MLiCl containing 1 μg/ml protease inhibitor (Sigma-Aldrich, St Louis,MO, USA) and incubated for 60 min at 37 °C. After centrifugation at10,000×g for 10 min at 4 °C, the supernatants were collected anddialysed against distilled water for 48 h at 4 °C using cellulosemembranes (10,000 Da cut-off) (Sigma-Aldrich, St Louis, MO, USA).Finally, the dialysates were lyophilized and their protein concentra-tion was determined by using the Bradford assay (Bio-Rad Laborato-ries, Inc., CA, USA). For comparison, whole-cell protein extracts fromlactobacilli were also obtained. Briefly, lactobacilli cells harvested bycentrifugationwere disrupted by homogenization using glass beads ina cell mill for 60s. The extracts were suspended in 1 ml of 0.01 M PBS,pH 7.0 and both the glass beads and the cell debris were removed bycentrifugation at 12,000×g for 10 min at 4 °C. The supernatantsobtained were designated as whole-cell protein extracts.

S-layer protein preparations as well as whole-cell protein extractswere subjected to SDS-PAGE (Laemmli, 1970) on vertical slab gelsusing a SE 250 Mighty Small (Hoefer, Inc. Holliston, MA, USA)electrophoresis apparatus. Protein bands were visualized by stainingthe gels with Coomassie brilliant blue and the prestained proteinmarker, broad range (6–175 kDa; New England Biolabs Inc. Hitchin,Hertfordshire, UK) was used as size marker.

2.9. Animals

Six- to 8-week-old BALB/c inbred mice with a body weight atabout 20–25 g and Fisher-344 inbred rats (130–180g) were kept incages in an environmentally controlled room with a 12 h light/darkcycle. Animals had free access to tap water and standard commercialfood pellets and experiments were performed in an accreditedestablishment (number EL 54 BIO 02, School of Biology, AristotleUniversity of Thessaloniki) according to both Greek National andEuropean Union legislation on animal experimentation. Rats wereonly used for the phagocytic experiments while each experimentalgroup of mice consisted of 5 animals and each experiment wasrepeated at least two times.

2.10. Air pouch formation and determination of exudate PMN cells

Air pouch formation on the dorsum ofmice or ratswas performed aspreviously described (Sedgwick et al., 1985) by the subcutaneousinjectionof 3 or 20 ml sterile air, respectively. Toachieve equal size, eachair pouch received a day-by-day appropriate air dose for 5 days.Administration of lactobacilli in the mouse or rat dorsal air pouch wasperformed on day 6 by the injection of 5×108CFU of bacteriaresuspended in 0.2 or 1 ml pyrogenic free saline, respectively. Controlanimals received an intra-air pouch administration of only pyrogenic

Fig. 1. Dendrogram of the genetic similarity between SfiI-digested DNA patterns oflactobacilli examined in this study. Cluster analysis has been performed with thePearson product moment correlation coefficient and by the unweighted pair groupalgorithm method with arithmetic averages (UPGMA).


free saline. Three hours after Lactobacillus injection, air pouch exudateswere collected and PMN cells were separated by centrifugation at300×g for 10 min. The cell pellets were washed twice by centrifugationin saline and PMN cells were counted using a haemocytometer whiletheir viability was assessed by trypan blue exclusion. The supernatantswerefiltered through0.22filters and stored at−20 °Cuntil further usedfor cytokine detection.

2.11. Phagocytic activity of air pouch PMN cells

Phagocytic activity of air pouch PMN cells was performed as wepreviously described (Kourelis et al., 2010c). The total number of PMNcells accumulated in the air pouch of individual mice was below thenumber of cells (2.5×106) required to perform the phagocytic assay.Thus, to avoid mixing air pouch cells from different mice we usedPMNs isolated from rat air pouch.

Cell suspension containing about 106 heat-killed yeast cells(Saccharomyces cerevisiae) was opsonized in rat serum for 30 min at37 °C. After opsonization, yeast cells were mixed with 2.5×106 airpouch PMNs and phagocytosis was performed for 30 min at 37 °C.Samples were taken at the indicated time post intervals of 5, 15 and30 min and were mixed with trypan blue. Phagocytosis of yeasts wasdetermined using haemocytometer under light microscope andphagocytic activity was expressed as percentage of PMNs containingbaker's yeast cells.

2.12. Determination of cytokines in air pouch exudates

Air pouch exudates isolated from control group or Lactobacillustreated air pouches were subjected to ELISA immunosorbent assay todetect the levels of gamma interferon (IFNγ), tumor necrosis factoralpha (TNFα) and interleukin (IL)-10 using the ELISA kit (eBioscience,Inc. San Diego, USA), following the manufacturer's protocol. Theresults were calculated against standard curves generated usingknown amounts of purified cytokines included in the kits. As positivesamples we also used air pouch exudates isolated from bacteriallipopolysaccharide (LPS) or Freund's complete adjuvant (FCA) treatedair pouches.

2.13. Oral administration of lactobacilli

For oral administration, mice were fed daily for 10 consecutivedays a dose of 109CFU of each of the Lactobacillus isolates resuspendedin 50 μl pyrogenic free saline. The control group received onlypyrogenic free saline given under the same conditions as those usedfor the test groups. One day after the last feeding the mice weresacrificed and the small intestine was removed, rinsed in saline andprocessed for immunohistochemical analyses.

2.14. Immunohistochemical detection of IgA, cytokine and TLR producingcells

Immunohistochemistry was performed in small intestine sectionsobtained from control and lactobacilli-orally fed mice. Small intestinepieces were fixed, dehydrated and finally embedded in paraffin blocksfollowing Sainte-Marie's (1962) technique. Paraffin inclused tissueswere cut in 4 µm sections and used for immunohistochemistryanalyses. The number of IgA and TLR2 producing cells was determinedby a direct immunofluorescence assay using conjugated FITCmonospecific anti-IgA (Sigma, St Louis, MO, USA) and TLR2 (SantaCruz Biotechnology, Inc., Santa Cruz, CA, USA) antibodies, respective-ly. The determination of TLRs 4, 9, cyclooxygenase (COX) and cytokineproducing cells was assayed by indirect immunofluorescence. Samplesections after deparaffinization and rehydration were blocked byincubation with 5% (v/v) FBS in PBS for 30 min at room temperature.After washing in PBS, sections were incubated for 2 h at room

temperature with properly diluted specific antibodies directed againstmouse TLRs 4, 9 (Abcam plc, Cambridge, UK), COX-1, COX-2, IL-5, IL-6,IL-12B (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), IFNγ,TNFα, IL-10 (R&D systems, Inc., Minneapolis, MN, USA). After 4washes in PBS the sections were incubated with anti-goat or anti-rabbit (for IL-5) FITC conjugated antibodies (Sigma, St Louis, MO, USA)for 1 h at room temperature, washed in PBS and examined by usingfluorescence light microscope. The number of fluorescent cells wascounted in 10 fields at 400× magnification and for the final scoringtwo investigators achieved consensus.

2.15. Statistical analysis

Evaluation of experimental data was performed as previouslydescribed (Vinderola et al., 2005) and results were expressed as meanvalues±standard error of the mean (SEM). Multiple comparisonswere performed by one-way analysis of variance (ANOVA) followedby Tukey's test. Data were considered significantly different when theP values were less than 0.05. All Lactobacillus properties examined inthis studywere analyzed by principal component analysis (PCA) usingthe Minitab software (Release 15, trial version; Minitab Inc.) aspreviously described (Giaouris et al., 2009).

3. Results

3.1. PFGE analysis

PFGE is considered as the gold standard method to discriminateprobiotics at strain level (FAO/WHO, 2002). To further characterizethe 12 Lactobacillus isolates included in this study, SfI digestedpatterns of their genomic DNA were subjected to PFGE analysis. Thereproducibility between different digests and electrophoretic runs forgenomes of individual strains was ≥97% and restriction patterns withgenetic homology above this percentage were considered identical.Cluster analysis of the PFGE macrorestriction patterns revealed thatthe DC421 and DC423 strains are genetically identical and conse-quently the 12 Lactobacillus isolates discriminated separately in 11different genotypes (Fig. 1). At a similarity level of 84% or above, allthe lactobacilli with the exception of L. plantarum 2012 strain weregrouped into 5 main clusters that correspond to the groupingpreviously obtained by SDS-PAGE of cell-free extracts isolated fromthese strains (Xanthopoulos et al., 1999). This grouping was furtherconfirmed by PFGE analysis of SmaI-digested genomic DNA (data notshown).


3.2. Adhesive capacity and cell surface traits of lactobacilli

The results presented in Table 1 show significant differences inhydrophobicity, autoaggregation and adhesion ability to Caco-2 cellsamong the Lactobacillus isolates tested, even within the same species.The DC421 and 2035 strains exhibit the highest hydrophobicity valuesthat is associated with increased autoaggregation ability andadherence capacity to Caco-2 monolayer (Fig. 2A–C). On the otherhand, strains that exhibit low percentage of cell surface hydropho-bicity also possess narrow range of low autoaggregation values andadhesion capacity. Interestingly, despite DC423 strain exhibits PFGEprofile that is similar with that of DC421, possess low hydrophobicityand autoaggregation as well as adhesive capacity. The above dataindicate potential correlation of the hydrophobicity with autoaggre-gation and adhesion to Caco-2 cell line.

Further analysis of the Lactobacillus cell surface traits by SDS-PAGEelectrophoresis revealed that only the probiotic L. acidophilus NCFB1748 strain possessed a potential S-layer protein. Fig. 2D shows thatthe L. acidophilus NCFB 1748 whole-cell protein sample (lane 1)contains a dominant band with an apparent molecular weight of45 kDa that is removed by the dissociating agent (lane 2) and finallyrecovered in the dialysed extract (lane 3). Furthermore, examinationof the Lactobacillus isolates for their capacity to produce EPS revealed

Fig. 2. Determination of adhesive capacity and S-layer proteins of lactobacilli. Adherence obacteria on Caco-2 cells (arrows) and on polysterene (arrowheads). (D) Representative SDproteins from L. acidophilus NCFB 1748 (lane 1), DC411 (lane 4), and DC422 (lane 7). Proteinafter treatment with 5 M LiCl. Protein profiles of dialysed extracts from L. acidophilus NCFB

that only the L. plantarum 2012 strain was identified as EPS-producingstrain (Table 1).

The above data show that only the Lactobacillus strains DC421,2035 and 2012 exhibit cell surface traits and/or high adhesive capacitysuggesting potential immunomodulatory activity.

3.3. Determination of the immunostimulatory activity of lactobacilliusing the dorsal air pouch system

To examine whether the lactobacilli included in this study elicitimmunostimulatory activity we used the air pouch system. The datapresented in Fig. 3 show that injection of lactobacilli into the air pouchof mice resulted in differential recruitment of PMN cells. Administra-tion of DC411, DC415, DC421, DC423, DC424, DC429 and 2035 strainsin the air pouch resulted in a subsequent significant (Pb0.05) increaseof the number of PMN cells in comparison with the control group.However, only the administration of DC421, DC429 and 2035 strainsresulted in accumulation of high number of PMNs that is statisticallysimilar to the number observed after administration of the probioticNCFB 1748 strain.

The strains DC421, DC429 and 2035 exhibiting increased chemo-taxis of PMNs in the air pouch were further examined for their abilityto activate the phagocytic capacity of PMNs as well as cytokine

f (A) DC423, (B) DC421 and (C) 2035 Lactobacillus strains to Caco-2 cells. Adhesion ofS-PAGE of the cell protein content of three different Lactobacillus strains. Whole-cells extracted from L. acidophilus NCFB 1748 (lane 2), DC411 (lane 5), and DC422 (lane 8),1748 (lane 3), DC411 (lane 6), and DC422 (lane 9).

Fig. 3. Induction of PMN cell recruitment in mouse air pouches following the injection ofdifferent lactobacilli. Each histogram bar represents the mean values of 5 mice±SEM. (*)Significantly different from the control group, Pb0.05; (#) Significant different incomparison with the probiotic NCFB 1748 treated group, Pb0.05.


production. The L. reuteri DC423 exhibiting low chemotactic activityand the NCFB 1748 strains were used as negative and positivecontrols, respectively. The data presented in Table 2 show that thestrains DC421, DC429 and 2035 as well as NCFB 1748 exhibit equalphagocytic activity that is significantly different than the control andthe DC423 strain.

Determination of cytokine levels in the air pouch exudates showedthat TNFα values were significantly increased (Pb0.05) for the DC423,DC421, DC429 and NCFB 1748 strains, while the strain 2035 inducedlow TNFα production in the air pouch. Furthermore, all strains wereobserved to induce low IFNγ production (Table 2). The administrationof DC421, DC429 and 2035 strains increased significantly the IL-10levels in comparison with the control group (Pb0.05).

All the above data indicate that the strains DC421, DC429 and 2035exhibit strong immunostimulatory activity in the air pouch systemsuggesting potential immunomodulatory activity in the gut mucosa.

3.4. Innate and adaptive immune responses in the gut mucosa of micefed with lactobacilli

We further examined the effect of DC421, DC429 and 2035 strainson the gutmucosa immune system following their oral administrationin healthy mice. NCFB 1748 and DC423 strains were also included aspositive and negative controls, respectively. Oral administration ofDC421, DC429 and 2035 in mice and subsequent haematoxylin–eosin stained sections of the small intestine showed no evidence ofinflammatory immune response or modification in the structure ofsmall intestine villi (Fig. 4A–B). Furthermore, the number of the small

Table 2Determination of phagocytic activity of PMN cells and cytokine production in Lactobacillus

Strains Phagocytic activity (%)a

5 min 15 min 30 min

Control 13.3±0.8 24.3±0.7 29.3±0.9NCFB 1748 30.6±3.2⁎ 50.7±2⁎ 68.0±0.6DC423 20.0±0.6# 38.7±0. 9⁎# 44.7±0.9DC421 43.3±2.3⁎# 54.3±0.9⁎ 69.0±5.7DC429 22.4±1.8⁎ 44.0±2.0⁎# 64.0±1.52035 36.0±0.6⁎ 65.0±1.7⁎# 73.7±1.8

a Mean percentage±SEM of air pouch cells with phagocytosed baker's yeasts (n=5).⁎ Significantly different from the control group, Pb0.05.# Significantly different from the probiotic NCFB 1748, Pb0.05.

intestine cells producing the COX-1 or COX-2 inflammatorymediatorswas similar in both the control and treated mice (Fig. 4C). Directimmunofluorescence analysis revealed that all strains, except forDC423, increased significantly the number of IgA positive cells incomparison with the control group (Fig. 4C).

Oral administration of NCFB 1748 increased the number of TNFαproducing cells in the lamina propria of small intestine (Fig. 5). Incontrast, the DC421, DC429 and 2035 strains decreased the number ofTNFα producing small intestine cells in comparison with the controlgroup and DC423 fed mice. Lactobacillus DC421, DC429 and 2035strains as well as the probiotic NCFB 1748 strain induced significantincrease (Pb0.05) in the number of IL-5, IL-6, IL-10 and IFNγproducing cells in comparison with the control group while did notaffect the number of IL-12 producing cells (Fig. 5). The above datademonstrate that orally administered lactobacilli regulate the secre-tion of critical immune mediators in the small intestine, in a strain-dependent manner and in the absence of severe inflammatoryresponse.

We examined whether the species specific cytokine profileinduced in the small intestine is associated with distinctive changesin pattern recognition receptors such as TLRs. All strains includingNCFB 1748 significantly increased the number of TLR2 and TLR4positive cells in comparison with the control group. In addition,increased number of TLR9 producing cells was observed in the smallintestine of the DC421, DC429 and NCFB 1748 fed mice (Fig. 6). Takentogether, the above data suggest that in response to DC421, DC429and 2035 oral administration, certain TLR signalling pathways areinvolved in the regulation of cytokine production and subsequentactivation of immune responses in the gut mucosa.

3.5. Distinction of strains exhibiting relevant probiotic characteristics byPCA

To discriminate the lactobacilli included in this study according totheir overall characteristics, we carried out PCA based on their cellsurface traits as well as their adhesive and immunomodulatorycapacity. The PCA results presented in Fig. 7, show that twocomponents which account for 86.1% of the variability of the originaldata set have been extracted. The first principal component thatdisplays 71.3% of the total variation clearly classifies the DC421,DC429 and 2035 Lactobacillus strains in a separate group, togetherwith the probiotic NCFB1748, suggesting that these strains may sharerelevant probiotic characteristics.

4. Discussion

The maintenance of gastrointestinal homeostasis is consideredcritical for preventing the development of immune-mediated andmetabolic-related diseases (Hotamisligil and Erbay, 2008). The positiveinfluence of probiotics on gut homeostasis is achieved by bacterialantagonism and immunomodulation, which help the healthy host tomaintain a “physiological state of inflammation” or to control several

treated mouse air pouches.

Concentration (pg/ml) of cytokines in mouse air pouch exudates

TNFα IFNγ IL-10

32.9±1.54 69.2±3.5 51.2±3.0⁎ 193.0±25.0⁎ 80.7±4.8 80.1±10.0⁎# 206.7±33.2⁎ 79.0±6.1 59.6±4.9⁎ 220.4±39.7⁎ 72.1±2.8 180.0±35.1⁎⁎ 344.7±37.2⁎# 87.0±2.5 163.3±12.0⁎⁎ 87.7±16.8 72.5±6.6 150.0±17.3⁎

Fig. 4. Effect of orally administered lactobacilli on the inflammatory or immune responsesof the intestinalmucosa ofmice. (A) Lightmicrophotographof haematoxylin–eosin stainedhistological sections from the small intestine of unfed mice. (B) Light microphotograph ofhaematoxylin–eosin stained histological sections from the small intestine ofmice receiving10 days of L. reuteri DC421 administration. (C) Determination of COX-1, COX-2 and IgAproducing cells in the lamina propria of the small intestine ofmice by immunofluorescencetest. Each histogram bar represents the mean values of 5 mice±SEM. Within each group,bars with different superscripts (a, b, c) differ significantly Pb0.05.

Fig. 5. Effect of orally administered lactobacilli on cytokine patterns of the gut mucosalimmune system of mice. Cytokines were determined on histological slices from smallintestine by immunofluorescence microscopy. Each histogram bar represents the meanvalues of 5 mice±SEM. Within each cytokine, bars with different superscripts (a, b, c)differ significantly, Pb0.05.

Fig. 6. Modulation of TLR expression profile in the small intestine of mice which wereorally fed with different lactobacilli. TLR positive cells were determined on histologicalslices from small intestine by immunofluorescence test. Each histogram bar representsthe mean values of 5 mice±SEM. Within each TLR, bars with different superscripts(a, b, c) differ significantly, Pb0.05.


infectious, inflammatory and immunologic reactions (Galdeano et al.,2007). However, given the recent concerns on probiotic treatment it isimportant to investigate their local and systemic immune activity

(Vogel, 2008). In this study we investigated the immunoregulatorycapacity of 12 presumptive probiotic lactobacilli previously isolatedfrom Feta cheese and infants' gastrointestinal tract and our analysis(Fig. 7) further distinguished three of them as potential candidateprobiotics.

Characterization of a strain and its origin are very important, asthis will provide an indication of its presumed safety and meetrequirements for exact information on the nomenclature of the strain(Salminen et al., 2001). PFGE analysis of genomic DNA isolated fromthe 12 lactobacilli included in this study confirmed their previousclassification (Xanthopoulos et al., 1999). Despite the fact that DC421and DC423 strains exhibit the same PFGE genomic DNA profile(Fig. 1), they express different cell surface traits as well as adhesivecapacity (Table 1) and immunoregulatory activity suggesting thatthey are distinct. Furthermore, these findingsmay indicate that DC423strain harbors mutation/s in genes directly or indirectly affecting itsimmunoregulatory activity and provide us with a promising strain forfurther comparative studies.

In our studywe used the air pouch system for initial discriminationof strains exhibiting immunostimulatory activity. Three of thesestrains, DC421, DC429 and 2035 exhibit immunostimulatory activity

Fig. 7. Principal component analysis based on bacterial cell surface traits as well as adhesive and immunomodulatory capacity data of the 12 Lactobacillus isolates and NCFB 1748strain. The first and second principal component display 71.3% and 14.8% of the total variation, respectively.


in the dorsal air pouch aswell as in the gut associated lymphoid tissue.Administration of these strains into the air pouch resulted in a rapidinflux of PMN cells possessing increased phagocytic activity andcytokine production. This immunostimulatory activity was of thesame potency with the activity induced by the probiotic L. acidophilusNCFB 1748 strain. Furthermore, oral administration of these strainsresulted in a strain specific production of a certain cytokine profile andTLR expression. Importantly, the environment is significant for acertain probiotic microorganism to exert its strain specific immuneactivity or cell interactions in the host (Lebeer et al., 2010). Theenvironment in the dorsal air pouch system and the intestine isdifferent. However, the air pouch lining tissue and the intestinemucosa share similarities since they both exhibit a barrier functionthat is achieved by epithelial cells and resident macrophages. Recentreports (Yam et al., 2008; Kourelis et al., 2010a, c) and the datapresented in this study indicate that Lactobacillus or yeast strainsinteract with the cells forming the barriers and initiate proinflamma-tory reactions such as the production of species specific chemokinesand cytokines as well as phagocytosis. Based on the above, the airpouch model provides the environment able to determine immuneresponses induced by lactobacilli or yeasts which resemble therespective responses in the small intestine and thus may be used asan alternative and rapid method for the initial discrimination andselection of potential presumptive probiotic Lactobacillus or yeaststrains.

The route of probiotic delivery and the subsequent early initiationof immune reactivity such as PMN cell recruitment and phagocyticactivity as well as chemokine and cytokine production at the local siteof administration are early events and important parameters involvedin non-well defined immune mechanisms. We previously showedthat the L. paracasei DC412 as well as probiotic L. acidophilus NCFB1748 strains exhibited differential interaction with the cells formingthe air pouch lining tissue with subsequent activation of a differentcytokine profile (Kourelis et al., 2010c). Based on the aboveobservations, we propose that the DC421, DC429 and 2035 strainspotentially also interact with the epithelial cells forming the air pouchlining tissue that in turn activate chemokine mediated chemotaxisand cytokine production. It has been shown that species specificrelease of a certain chemokine profile is induced in the air pouch inresponse to Lactococcus injection (Yam et al., 2008). In addition,Lactococcus lactis administration in the air pouch increased PMNrecruitment while PMN recruitment in the udder due to L. lactisadministration successfully treated bovine mastitis (Crispie et al.,

2008) suggesting that DC421, DC429, 2035 and NCFB 1748 strainsmay also exhibit pathogenic bacterial clearance activity. We proposethat these early proinflammatory responses in the air pouch sharesimilarities with the responses induced by these strains in theintestine after their oral delivery.

DC421, DC429 and 2035 strains exhibited in vitro probioticproperties such as acid and bile salt tolerance indicating their capacityto survive in the intestinal tract (Xanthopoulos et al., 2000b).Characterization of several Lactobacillus strains as presumptiveprobiotics is based on their ability to persist in the intestine throughtheir adhesive capacity. Hydrophobicity as well as autoaggregation ofbacteria is a phenotype related to their adhesive capacity (Del Re et al.,2000; Kos et al., 2003). The strains DC421 and 2035 exhibithydrophobicity, autoaggregation and adhesive capacity to Caco-2cell line indicating that their immmunoregulatory activity in the GALTmay be associated with their potential capacity to colonize the hostgut mucosa. Furthermore, DC421 and 2035 strains exhibited capacityto bind on polystyrene surface (Fig. 2A–C) indicating their potentialability to form biofilm (Jones and Versalovic, 2009). The strain DC423consistently showed low immunoregulatory activity in associationwith low range of hydrophobicity, autoaggregation, and no adhesivecapacity on Caco-2 cell line or polysterene surface. In contrast, theDC429 and NCFB 1748 strains possess strong immunostimulatoryactivity as well as low hydrophobicity and autoaggregation values inassociation with no adhesive ability on Caco-2 cell line. We havepreviously showed that FITC-labelled NCFB-1748 strain interacts withthe GALT (Kourelis et al., 2010c) after its oral delivery. Taken together,all the above data indicate that cell surface traits and adhesion abilitymay not be the only criteria but alternative and importantcomponents of a complex mechanism which enable a microorganismto bind and persist in the host gut to exert its immunoregulatoryactivity.

The potential probiotic Lactobacillus strains were also examinedfor the presence of S-layer proteins and the production of EPS(Table 1). EPS is produced only by the 2012 strain possessing lowimmunostimulatory activity in the air pouch. The S-layer proteinspresent on cell surface of several L. acidophilus strains (Smit et al.,2001; Frece et al., 2005b) were also detected in L. acidophilus NCFB1748 strain (Fig. 2D). A 45 kDa S-layer protein present on the surfaceof L. acidophilus NCFM strain was found to be involved in theregulation of immature dendritic cells as well as cytokine production(Konstantinov et al., 2008). We propose that the 45 kDa S-layerprotein of NCFB 1748 strain detected in this study may be involved in


the regulation of cytokine profile observed in the small intestine afteroral administration of this strain.

The interactivity of the DC421, DC429 and 2035 as well as NCFB 1748strains in the small intestine after their oral delivery is not followed byactivation of a general inflammatory response since we did not observeany change in the massive influx of PMNs (Fig. 3), in the structure ofthe villi, as well as in the number of cells producing the inflammatorymediators COX-1 and COX-2 (Fig. 4A–C). Oral delivery of DC421, DC429and 2035 as well as NCFB 1748 strains exhibited immunoenhancingactivity in the small intestine concerning innate and adaptive immuneresponses such as IgA (Fig. 4C) and cytokineproduction (Fig. 5). The strainDC423 exhibiting low immunoregulatory capacity in the air pouchprovided poor immune responses after its oral delivery, further support-ing the eligibility of the air pouch system to discriminate Lactobacillusstrains exhibiting immunostimulatory activity. The increase in IgAproduction may be due, at least in part, to the observed up-regulation ofIL-6 (Fig. 5) that promotes B-cell differentiation into antibody producingcells (Vinderola et al., 2005). Activation of the innate immunity and IgAproduction are thought to be the most important mechanism involved ingutmucosa immune stimulationbyprobiotics (Galdeanoet al., 2007). Theenhancement of the intestinal IgA response it has been associated withimprovement of the gastrointestinal resistance to pathogenic micro-organisms (Isolauri et al., 1994).

Our findings demonstrate that DC421 and DC429 strains activatethe cytokine profile of IL-5, IL-6, IL-10 and IFNγ through the increasedexpression of TLR2, TLR4 and TLR9 while the 2035 strain induce thesame cytokine profile through TLR2 and TLR4, indicating the existenceof alternative signalling pathways for cytokine production in the gutmucosa. These observations suggest that DC421, DC429 and 2035mayinduce a balanced Th1/Th2 response in the GALT. Our data clearlyshow that oral delivery of DC421, DC429 and 2035 in mice results indownregulation of TNFα production in the GALT (Fig. 5) while thatdifferentiates this response from the data showing activation of TNFαproduction by these strains in the air pouch (Table 2). Positive andnegative regulation of TNFα production in the air pouch and in theGALT respectively by L. reuteri DC421 and L. rhamnosus 2035 strainsmay be due to their potential capacity to form biofilm in the gutmucosa. Several presumptive probiotics were found to downregulateTNFα production in vitro (Zhang et al., 2005; Lin et al., 2008) whileonly biofilm forming L. reuteri strains inhibits the in vitro TNFαproduction by LPS-activated humanmonocytes (Jones and Versalovic,2009).

The involvement of TLRs in the mechanism that activates theinnate immune response in the gut mucosa is well documented(Takeda et al., 2003). TLR2 has been shown to be a signal transducerwhich recognizes peptidoglycan fragments and lipoteichoic acid fromGram-positive bacteria, while TLR4 is essential for the recognition oflipopolysaccharides (LPS) from Gram-negative bacteria (Vinderolaet al., 2005). Furthermore, TLR9 signalling activated by immunosti-mulatory bacterial CpG DNA sequences is essential in mediating anti-inflammatory effects of probiotic strains (Rachmilewitz et al., 2004).Taking into consideration that colonization of the gastrointestinaltract with lactobacilli interferes with colonization of enteropathogen-ic microorganisms (Shu and Gill, 2002), the TLR4-signalling inductionby DC421, DC429 and 2035 strains might result in health benefitsthrough competition with Gram-negative pathogens for adhesionsites (Vinderola et al., 2005).

In conclusion, the L. reuteri DC421, L. rhamnosus DC429 andL. plantarum 2035 strains exhibited regulatory activity of earlyimmune responses in both air pouch and intestine that is character-ized by stimulation of PMN chemotaxis, phagocytic activity, combi-nation of TLR2/TLR4/TLR9 signalling and secretion of a certaincytokine profile. The immunostimulatory activity of DC421 and2035 strains occurs in association with their adhesive capacity andcertain cell surface traits such as hydrophobicity and autoaggrega-tion. DC429 and NCFB 1748 strains activate immune responses in the

GALT by an alternative and unknown mechanism that may involvedirect activation of immune cells by Lactobacillus antigens. Theactivation of these specific immune responses by these nonpatho-genic microorganisms helps the achievement of homeostasis andprovides the healthy host with a higher capacity to resist anyinflammatory response. Final validation concerning the effectivenessof DC421, DC429 and 2035 strains must await further experimenta-tion using animal disease models and clinical trials.

References

Chauviere, G., Coconnier, M.H., Kerneis, S., Fourniat, J., Servin, A.L., 1992. Adhesion ofhuman Lactobacillus acidophilus strain LB to human enterocyte-like Caco-2 cells.Journal of General Microbiology 138, 1689–1696.

Coates, N.J., McColl, S.R., 2001. Production of chemokines in vivo in response tomicrobial stimulation. Journal of Immunology 166, 5176–5182.

Crispie, F., Alonso-Gomez, M., O'Loughlin, C., Klostermann, K., Flynn, J., Arkins, S.,Meaney, W., Paul Ross, R., Hill, C., 2008. Intramammary infusion of a live culture fortreatment of bovine mastitis: effect of live lactococci on the mammary immuneresponse. Journal of Dairy Research 75, 374–384.

Cross, M.I., Mortensen, R.I., Kudsk, J., Gill, H.I., 2002. Dietary intake of Lactobacillusrhamnosus HN001 enhances production of both Th1 and Th2 cytokines in antigen-primed mice. Medical Microbiology and Immunology 191, 49–53.

Crow, V., Gopal, P., Wicken, A., 1995. Cell surface differences of lactococcal strains.International Dairy Journal 5, 45–68.

Del Re, B., Sgorbati, B., Miglioli, M., Palenzona, D., 2000. Adhesion, autoaggregation andhydrophobicity of 13 strains of Bifidobacterium longum. Letters in AppliedMicrobiology 31, 438–442.

FAO/WHO, 2001. Health and Nutritional Properties of Probiotics in Food includingPowder Milk with Live Lactic Acid Bacteria. Report of a Joint FAO/WHO ExpertConsultation on Evaluation of Health and Nutritional Properties of Probiotics inFood Including Powder Milk with Live Lactic Acid Bacteria. Córdoba, Argentina.

FAO/WHO, 2002. Guidelines for the Evaluation of Probiotics in Food. Food andAgriculture Organization of the United Nations and World Health OrganizationWorking Group Report, London Ontario, Canada.

Frece, J., Kos, B., Beganovic, J., Vukovic, S., Suskovic, J., 2005a. In vivo testing of functionalproperties of three selected probiotic strains. World Journal of Microbiology andBiotechnology 21, 1401–1408.

Frece, J., Kos, B., Svetec, I., Zgaga, Z., Mrsa, V., Suskovic, J., 2005b. Importance of S-layerproteins in probiotic activity of Lactobacillus acidophilus M92. Journal of AppliedMicrobiology 98, 285–292.

Galdeano, C.M., de Moreno de LeBlanc, A., Vinderola, G., Bonet, M.E., Perdigon, G., 2007.Proposed model: mechanisms of immunomodulation induced by probioticbacteria. Clinical and Vaccine Immunology 14, 485–492.

Garrote, G., Delfederico, L., Bibiloni, R., Abraham, A., Perez, P., Semorile, L., Antoni, G.,2004. Lactobacilli isolated from kefir grains: evidence of the presence of S-layerproteins. Journal of Dairy Research 71, 222–230.

Giaouris, E., Chapot-Chartier, M.P., Briandet, R., 2009. Surface physicochemical analysisof natural Lactococcus lactis strains reveals the existence of hydrophobic and lowcharged strains with altered adhesive properties. International Journal of FoodMicrobiology 131, 2–9.

Hotamisligil, G.S., Erbay, E., 2008. Nutrient sensing and inflammation in metabolicdiseases. Nature Reviews Immunology 8, 923–934.

Isolauri, E., Kaila, M., Mykkanen, H., Ling, W.H., Salminen, S., 1994. Oral bacteriotherapyfor viral gastroenteritis. Digestive Diseases and Sciences 39, 2595–2600.

Jacobsen, C., Nielsen, V., Hayford, A., Moller, L., Michaelsen, K., Parregaard, A.,Sandstrom, B., Tvede, M., Jakobsen, M., 1999. Screening of probiotic activities offorty-seven strains of Lactobacillus spp. by in vitro techniques and evaluation of thecolonization ability of five selected strains in humans. Applied and EnvironmentalMicrobiology 65, 4949–4956.

Jones, S.E., Versalovic, J., 2009. Probiotic Lactobacillus reuteri biofilms produceantimicrobial and anti-inflammatory factors. BMC Microbiology 9, 35.

Konstantinov, S.R., Smidt, H., de Vos, W.M., Bruijns, S.C., Singh, S.K., Valence, F., Molle,D., Lortal, S., Altermann, E., Klaenhammer, T.R., van Kooyk, Y., 2008. S layer proteinA of Lactobacillus acidophilus NCFM regulates immature dendritic cell and T cellfunctions. Proceedings of the National Academy of Sciences of the United States ofAmerica 105, 19474–19479.

Kos, B., Suskovic, J., Vukovic, S., Simpraga, M., Frece, J., Matosic, S., 2003. Adhesion andaggregation ability of probiotic strain Lactobacillus acidophilus M92. Journal ofApplied Microbiology 94, 981–987.

Kotzamanidis, C., Zdragas, A., Kourelis, A., Moraitou, E., Papa, A., Yiantzi, V., Pantelidou,C., Yiangou, M., 2009. Characterization of vanA-type Enterococcus faecium isolatesfrom urban and hospital wastewater and pigs. Journal of Applied Microbiology 107,997–1005.

Kourelis, A., Kotzamanidis, C., Litopoulou-Tzanetaki, E., Papaconstantinou, J., Tzanetakis,N., Yiangou, M., 2010a. Immunostimulatory activity of potential probiotic yeaststrains in the dorsal air pouch system and the gut mucosa. Journal of AppliedMicrobiology. doi:10.1111/j.1365-2672.2009.04651.x.

Kourelis, A., Kotzamanidis, C., Litopoulou-Tzanetaki, E., Scouras, Z., Tzanetakis, N.,Yiangou, M., 2010b. Determination of in vitro probiotic properties of yeast strainsisolated from feta cheese and infants. Journal of Biological Research 13, 93–104.

Kourelis, A., Zinonos, I., Kakagianni,M., Christidou, A., Christoglou, N., Yiannaki, E., Testa, T.,Kotzamanidis, C., Litopoulou-Tzanetaki, E., Tzanetakis, N., Yiangou, M., 2010c.


Validation of the dorsal air pouch model to predict and examine immunostimulatoryresponses in the gut. Journal of Applied Microbiology 108, 274–284.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature 227, 680–685.

Lebeer, S., Vanderleyden, J., De Keersmaecker, C.J.S., 2010. Host interactions of probioticbacterial surface molecules: comparison with commensals and pathogens. NatureReviews Microbiology 8, 171–184.

Lin, Y.P., Thibodeaux, C.H., Pena, J.A., Ferry, G.D., Versalovic, J., 2008. ProbioticLactobacillus reuteri suppress proinflammatory cytokines via c-Jun. InflammatoryBowel Diseases 14, 1068–1083.

Mohamadzadeh, M., Olson, S., Kalina, W.V., Ruthel, G., Demmin, G.L., Warfield, K.L.,Bavari, S., Klaenhammer, T.R., 2005. Lactobacilli activate human dendritic cells thatskew T cells toward T helper 1 polarization. Proceeding of the National academy ofSciences of the United States of America 102, 2880–2885.

Neish, J.M.T., 2002. The gut microflora and intestinal epithelial cells: a continuingdialogue. Microbes and Infection 4, 309–317.

Nerstedt, A., Nilsson, E.C., Ohlson, K., Hakansson, J., Thomas Svensson, L., Lowenadler, B.,Svensson, U.K., Mahlapuu, M., 2007. Administration of Lactobacillus evokes coordinatedchanges in the intestinal expression profile of genes regulating energy homeostasis andimmune phenotype in mice. British Journal of Nutrition 97, 1117–1127.

Ouwehand, A.C., Kirjavainen, P.V., Shortt, C., Salminen, S., 1999. Probiotics: mechanismsand established effects. International Dairy Journal 9, 43–52.

Pelletier, C., Bouley, C., Cayuela, C., Bouttier, S., Bourlioux, P., Fontaine, M., 1997. Cellsurface characteristics of Lactobacillus casei subsp. casei, Lactobacillus paracaseisubsp. paracasei, and Lactobacillus rhamnosus strains. Applied and EnvironmentalMicrobiology 63, 1725–1731.

Puertollano, E., Puertollano, M.A., Cruz-Chamorro, L., Alvarez de Cienfuegos, G., Ruiz-Bravo, A., De Pablo, M.A., 2008. Orally administered Lactobacillus plantarum reducespro-inflammatory interleukin secretion in sera from Listeria monocytogenesinfected mice. British Journal of Nutrition 99, 819–825.

Rachmilewitz, D., Katakura, K., Karmeli, F., Hyashi, T., Reinus, C., Rudensky, B., Akira, S.,Takeda, K., Lee, J., Takabayashi, K., Raz, E., 2004. Toll-like receptor 9 signalingmediatesthe anti-inflammatory effects of probiotics in murine experimental colitis. Gastroen-terology 126, 520–528.

Sainte-Marie, G., 1962. A paraffin embedding technique for studies employingimmunofluorescence. Journal of Histochemistry and Cytochemistry 10, 250–256.

Salminen, S.J., von Wright, A.J., Ouwehand, A.C.H., Holzapfel, W.H., 2001. Safetyassessment of probiotics and starters. In: Adams, M.R., Nout, M.J.R. (Eds.),Fermentation and Food Safety. Aspen Publishers, Gaithersburg, pp. 239–251.

Sedgwick, A.D., Moore, A.R., Al-Duaij, A.Y., Edwards, J.C., Willoughby, D.A., 1985. Theimmune response to pertussis in the 6-day air pouch: a model of chronic synovitis.British Journal of Experimental Pathology 66, 455–464.

Sengul, N., Aslim, B., Ucar, G., Yucel, N., Isik, S., Bozkurt, H., Sakaogullari, Z., Atalay, F.,2006. Effects of exopolysaccharide-producing probiotic strains on experimentalcolitis in rats. Diseases of the Colon and Rectum 49, 250–258.

Shu, Q., Gill, H.S., 2002. Immune protection mediated by the probiotic Lactobacillusrhamnosus HN001 (DR20) against Escherichia coli O157:H7 infection in mice. FEMSImmunology and Medical Microbiology 34, 59–64.

Smit, E., Oling, F., Demel, R., Martinez, B., Pouwels, P.H., 2001. The S-layer protein ofLactobacillus acidophilus ATCC 4356: identification and characterisation of domainsresponsible for S-protein assembly and cell wall binding. Journal of MolecularBiology 305, 245–257.

Takeda, K., Kaisho, T., Akira, S., 2003. Toll-like receptors. Annual Reviews in Immunology21, 335–376.

Tzanetakis, N., Litopoulou-Tzanetaki, E., 1992. Changes in numbers and kinds of lacticacid bacteria in Feta and Teleme, two Greek cheeses from ewes' milk. Journal ofDairy Science 75, 1389–1393.

Vaarala, O., 2003. Immunological effects of probiotics with special reference tolactobacilli. Clinical and Experimental Allergy 33, 1634–1640.

Van Geel-Schutte, G., Flesch, F., ten Brink, B., Smith, M., Dijkhuizen, L., 1998. Screeningand characterization of Lactobacillus strains producing large amounts of exopoly-saccharides. Applied Microbiology and Biotechnology 50, 697–703.

Vinderola, G., Matar, C., Perdigon, G., 2005. Role of intestinal epithelial cells in immuneeffects mediated by Gram-positive probiotic bacteria: involvement of toll-likereceptors. Clinical and Diagnostic Laboratory Immunology 12, 1075–1084.

Vogel, G., 2008. Clinical trials. Deaths prompt a review of experimental probiotictherapy. Science 319, 557.

Voltan, S., Castagliuolo, I., Elli,M., Longo, S., Brun, P., D'Inca, R., Porzionato, A.,Macchi, V., Palu,G., Sturniolo, G.C., Morelli, L., Martines, D., 2007. Aggregating phenotype in Lactobacilluscrispatus determines intestinal colonization and TLR2 and TLR4 modulation in murinecolonic mucosa. Clinical and Vaccine Immunology 14, 1138–1148.

Wittig, B.M., Zeitz, M., 2003. The gut as an organ of immunology. International Journalof Colorectal Disease 18, 181–187.

Xanthopoulos, V., Ztaliou, I., Gaier, W., Tzanetakis, N., Litopoulou-Tzanetaki, E., 1999.Differentiation of Lactobacillus isolates from infant faeces by SDS-PAGE and rRNA-targeted oligonucleotide probes. Journal of Applied Microbiology 87, 743–749.

Xanthopoulos, V., Hatzikamari, M., Adamidis, T., Tsakalidou, E., Tzanetakis, N., Litopoulou-Tzanetaki, E., 2000a. Heterogeneity of Lactobacillus plantarum isolates from Fetacheese throughout ripening. Journal of Applied Microbiology 88, 1056–1064.

Xanthopoulos, V., Litopoulou-Tzanetaki, E., Tzanetakis, N., 2000b. Characterization ofLactobacillus isolates from infant faeces as dietary adjuncts. Food Microbiology 17,205–215.

Yam, K.K., Pouliot, P., N'Diaye, M.M., Fournier, S., Olivier, M., Cousineau, B., 2008. Innateinflammatory responses to the Gram-positive bacterium Lactococcus lactis. Vaccine26, 2689–2699.

Zdragas, A., Partheniou, P., Kotzamanidis, C., Psoni, L., Koutita, O., Moraitou, E., Tzanetakis,N., Yiangou, M., 2008. Molecular characterization of low-level vancomycin-resistantenterococci found in coastal water of Thermaikos Gulf, Northern Greece. WaterResearch 42, 1274–1280.

Zhang, L., Li, N., Caicedo, R., Neu, J., 2005. Alive and dead Lactobacillus rhamnosus GGdecrease tumor necrosis factor-alpha-induced interleukin-8 production in Caco-2cells. Journal of Nutrition 135, 1752–1756.